Dearth of Data Science talent in India. Why?
I was recently invited as an interviewer at a mock interview event at Upgrad. I interviewed fresh college grads enrolled in a 5 month Data Science certification programme. Out of the 10 students i interviewed, only 2 made it through.
The objective was to fill a junior data science position, which according to me is a low expectations job. where one is selected primarily on potential they display rather than possessing real industry required skills.
This observation of a lack of employable candidates in data science resonates with my past experiences of forming data science teams in India. The success rate, which i am sure many others like me would concur, for successful data science hires is less than 15%.
There simply isn’t nowhere near enough skilled personnel out there that can fill data science and analytics positions in India. This situation is ironic for, as many may concur, for every data science related job posting out there, you will find 100+ applications.
There are well documented reasons for a remarkable increase in the number of people wanting to get into data science in the past 5 years, much more so in the last 2. The question that i scratch my head upon is that there are several data science training schools out there, many of which have high quality content like coursera, udacity, dataquest.io (without dwelling into more unpaid promotion :) ). What is the reason that these schools are not able to produce highly skilled Data Science talent in India?
Upon taking a deep dive, multiple reasons unfold that can answer this question.
1. Quality courses out there are too difficult to comprehend for their targeted audience.
Websites like coursera, Udacity, dataquest.io (to name a few) provide excellent MOOCS for a range of subjects ranging from data science fundamentals to Self driving cars to Machine translation for chat-bots.
The feedback i got from my acquaintances learning Data analytics is that the High quality courses, the really good ones are too steep a climb for someone who is getting his/her first taste in data science.
Even though course creators make sure that course curriculum is simplified to dumb down the underlying math and stats concepts, It still is complicated to understand for someone who is studying a concept like Support Vector machine for the first time and apply it to solve real world problems.
2. Lack of direction in choosing one’s right career path in Analytics
The field of Data Analytics trickles down to very distinct sub disciplines. Data Scientist, Data Engineering, Analyst, Research Scientist. A few specialised ones are also prevalent like Computer vision engineer, NLP Engineer, Machine learning engineer. Each of these require very different and special skills for them to be industry relevant.
There seemed to be a lack of direction in most younger people I've interviewed or spoken to (experienced and freshmen) about the right career path for them to take in data science.
3. Absence of actionable business projects/use cases in the curriculum
This is what differentiates the good courses from the bad ones. Most of the courses out there lack actionable business projects/use cases, that resemble actual work in the industry, in the curriculum. Sadly, the bad ones out there are the ones that are easier for a beginner to comprehend.
4. Programs lack individual attention for the ones who need it
Most of the time, If you are teaching somebody a tricky subject or concept, individual attention is required to ensure they grasp the concept properly. Many programs employ mentors to tackle this specific problem, but it is safe to say that training programmes are failing to provide enough individual attention to students for them to be effective.
These observations hold true for short duration trainings such as bootcamps as well.
Not only is industry facing a shortage of good hirable people, I feel that there is also a shortage of skilled trainers out there.
The gap between industry and academia can only be filled if there are either more industry experts who want to teach, or more teachers that are willing to gain industry exposure by participating in exchange engagements.
The points highlighted should serve to bring some light to the situation and solve the talent problem. There are plenty of bright people out there.
The talent pool of intelligent individual is pretty large. If they get the right guidance and training, there is no reason why it can’t translate to forming a rich talent pool of Data Scientists and Analysts in India.